PHP tidy_repair_string函数代码示例

 /**
  * Repair HTML. If Tidy not exists, use repair function.
  *
  * @param   string  $html     The HTML string to repair.
  * @param   boolean $use_tidy Force tidy or not.
  *
  * @return  string  Repaired HTML.
  */
 public static function repair($html, $use_tidy = true)
 {
     if (function_exists('tidy_repair_string') && $use_tidy) {
         $TidyConfig = array('indent' => true, 'output-xhtml' => true, 'show-body-only' => true, 'wrap' => false);
         return tidy_repair_string($html, $TidyConfig, 'utf8');
     } else {
         $arr_single_tags = array('meta', 'img', 'br', 'link', 'area');
         // Put all opened tags into an array
         preg_match_all("#<([a-z]+)( .*)?(?!/)>#iU", $html, $result);
         $openedtags = $result[1];
         // Put all closed tags into an array
         preg_match_all("#</([a-z]+)>#iU", $html, $result);
         $closedtags = $result[1];
         $len_opened = count($openedtags);
         // All tags are closed
         if (count($closedtags) == $len_opened) {
             return $html;
         }
         $openedtags = array_reverse($openedtags);
         // Close tags
         for ($i = 0; $i < $len_opened; $i++) {
             if (!in_array($openedtags[$i], $closedtags)) {
                 if (!in_array($openedtags[$i], $arr_single_tags)) {
                     $html .= "</" . $openedtags[$i] . ">";
                 }
             } else {
                 unset($closedtags[array_search($openedtags[$i], $closedtags)]);
             }
         }
         return $html;
     }
 }

function autoclose_tags_custom($content = '')
{
    if (function_exists('tidy_repair_string')) {
        return tidy_repair_string($content, array('clean' => true, 'drop-font-tags' => true, 'drop-proprietary-attributes' => true, 'enclose-text' => true), 'utf8');
    }
    preg_match_all("#<([a-z]+)( .*)?(?!/)>#iU", $content, $result);
    $openedtags = $result[1];
    preg_match_all("#</([a-z]+)>#iU", $content, $result);
    $closedtags = $result[1];
    $len_opened = count($openedtags);
    if (count($closedtags) == $len_opened) {
        return $content;
    }
    $openedtags = array_reverse($openedtags);
    for ($i = 0; $i < $len_opened; $i++) {
        if (!in_array($openedtags[$i], $closedtags)) {
            if (!in_array($openedtags[$i], array('img', 'br', 'hr', 'input', 'col', 'meta', 'link'))) {
                $content .= '</' . $openedtags[$i] . '>';
            }
        } else {
            unset($closedtags[array_search($openedtags[$i], $closedtags)]);
        }
    }
    return $content;
}

function DoFooter($buffer)
{
    global $noFooter, $timeStart, $queries, $overallTidy, $boardname, $title, $dblink, $ajax, $footerButtons, $footerExtensionsA, $footerExtensionsB;
    if (!$noFooter) {
        //if(function_exists("runBucket")) runBucket("footerButtons");
        $footer = format("\n\t\t<div class=\"footer\">\n\t\t\tPowered by <a href=\"https://github.com/Dirbaio/ABXD\">AcmlmBoard XD</a>, version 2.2.6<br />\n\t\t\tBy Kawa, Mega-Mario, Nikolaj, et al<br />\n\t\t\tAcmlmBoard &copy; Jean-Fran&ccedil;ois Lapointe<br />\n\t\t\t" . __("Page rendered in {0} seconds with {1}.") . "<br />\n\t\t\t{3}\n\n\t\t\t<a href=\"http://validator.w3.org/check?uri=referer\">\n\t\t\t\t<img src=\"img/xhtml10.png\" alt=\"Valid XHTML 1.0 Transitional\" />\n\t\t\t</a>\n\t\t\t<a href=\"http://jigsaw.w3.org/css-validator/\">\n\t\t\t\t<img src=\"img/css.png\" alt=\"Valid CSS!\" />\n\t\t\t</a>\n\t\t\t<a href=\"http://abxd.dirbaio.net/?page=downloads\">\n\t\t\t\t<img src=\"img/getabxd.png\" alt=\"Get a copy for yourself\" />\n\t\t\t</a>\n\t\t\t{2}\n\t\t</div>\n\t</div>\n</body>\n</html>\n", sprintf("%1.3f", usectime() - $timeStart), Plural($queries, __("MySQL query")), $footerButtons, __("<!-- English translation by Kawa -->"));
    }
    $boardTitle = htmlval($boardname);
    if ($title != "") {
        $boardTitle .= " &raquo; " . $title;
    }
    $raw = $buffer . $footerExtensionsA . $footer . $footerExtensionsB;
    $raw = str_replace("<title>[[BOARD TITLE HERE]]</title>", "<title>" . $boardTitle . "</title>", $raw);
    if (!$ajax) {
        $raw = OptimizeLayouts($raw);
    }
    mysql_close($dblink);
    if (!$overallTidy) {
        return $raw;
    }
    $tidyConfig = array("show-body-only" => 0, "output-xhtml" => 1, "doctype" => "transitional", "logical-emphasis" => 1, "alt-text" => "", "drop-proprietary-attributes" => 1, "wrap" => 0, "preserve-entities" => 1, "indent" => 1, "input-encoding" => "utf8", "char-encoding" => "utf8", "output-encoding" => "utf8", "new-blocklevel-tags" => "video");
    //if(function_exists(OptimizeLayouts))
    //	$raw = OptimizeLayouts($raw);
    $clean = tidy_repair_string($raw, $tidyConfig);
    $clean = str_replace("class=\"required", "required=\"required\" class=\"", $clean);
    $textareaFixed = str_replace("\r", "", $clean);
    $textareaFixed = str_replace(" </text", "</text", $textareaFixed);
    $textareaFixed = str_replace("\n</text", "</text", $textareaFixed);
    //$textareaFixed = str_replace("\n</text", "</text", $textareaFixed);
    return $textareaFixed;
}

 public function flush()
 {
     if ($this->enabled && $this->userEnabled) {
         $this->outputResponse->appendData(\tidy_repair_string($this->data, $this->config, 'utf8'));
     }
     $this->outputResponse->flush();
     $dataBuffer = '';
 }

 function __destruct2()
 {
     $output = ob_get_clean();
     $config = array('indent' => true, 'output-xhtml' => false, 'wrap' => false, 'hide-comments' => true, 'logical-emphasis' => true);
     $output = tidy_repair_string($output, $config, 'utf8');
     ob_start();
     echo $output;
 }

function webifyFile($file, $toc, $editions)
{
    $filename = basename($file);
    if (strpos('-book-', $filename)) {
        return;
    }
    $toc = str_replace('<a href="' . $filename . '">', '<a href="' . $filename . '" class="active">', $toc);
    $template = file_get_contents(dirname(__FILE__) . DIRECTORY_SEPARATOR . 'templates' . DIRECTORY_SEPARATOR . 'page.html');
    $title = '';
    $content = '';
    $prev = '';
    $next = '';
    if ($filename !== 'index.html') {
        if (strpos($filename, 'appendixes') === 0) {
            $type = 'appendix';
        } else {
            if (strpos($filename, 'preface') === 0) {
                $type = 'preface';
            } else {
                if (strpos($filename, 'pt') === 0) {
                    $type = 'part';
                } else {
                    $type = 'chapter';
                }
            }
        }
        $buffer = file_get_contents($file);
        $title = getSubstring($buffer, '<title>', '</title>', FALSE, FALSE);
        $content = getSubstring($buffer, '<div class="' . $type . '"', '<div class="navfooter">', TRUE, FALSE);
        $prev = getSubstring($buffer, '<link rel="prev" href="', '" title', FALSE, FALSE);
        $next = getSubstring($buffer, '<link rel="next" href="', '" title', FALSE, FALSE);
        if (!empty($prev)) {
            $prev = '<a accesskey="p" href="' . $prev . '">Prev</a>';
        }
        if (!empty($next)) {
            $next = '<a accesskey="n" href="' . $next . '">Next</a>';
        }
    } else {
        $buffer = file_get_contents($file);
        $title = getSubstring($buffer, '<title>', '</title>', FALSE, FALSE);
        // $content = getSubstring($buffer, '<div class="titlepage"', '<div class="navfooter">', TRUE, FALSE);
        $content = getSubstring($buffer, '<div class="titlepage"', '<hr>', TRUE, FALSE) . "\n</div>\n";
        $prev = '';
        $next = getSubstring($buffer, '<link rel="next" href="', '" title', FALSE, FALSE);
        if (!empty($next)) {
            $next = '<a accesskey="n" href="' . $next . '">Next</a>';
        }
    }
    $buffer = str_replace(array('{title}', '{content}', '{toc}', '{editions}', '{prev}', '{next}'), array($title, $content, $toc, $editions, $prev, $next), $template);
    if (function_exists('tidy_repair_string')) {
        $buffer = tidy_repair_string($buffer, array('indent' => TRUE, 'output-xhtml' => TRUE, 'wrap' => 0), 'utf8');
    }
    file_put_contents($file, $buffer);
}

 function TidyLayout(&$header, &$footer)
 {
     return;
     global $tidyconfig;
     print "<!-- TIDYLAYOUT \n" . $header . "\n\n" . $footer;
     $sep = "%%SNIP%%";
     $pl = trim(tidy_repair_string($header . $sep . $footer, $tidyconfig));
     $header = substr($pl, 0, strpos($pl, $sep));
     $footer = substr($pl, strpos($pl, $sep) + strlen($sep));
     print "\n\n" . $header . "\n\n" . $footer . "\n-->";
 }

 private function parse($data)
 {
     $data = iconv('gbk', 'utf-8', $data);
     $data = tidy_repair_string($data);
     $doc = phpQuery::newDocument($data);
     foreach (pq('tr') as $index => $tr) {
         if ($index < 3) {
             continue;
         }
         $item = array('date' => pq($tr)->find('td:eq(0)')->text(), 'tag' => pq($tr)->find('td:eq(2)')->text(), 'number' => pq($tr)->find('td:eq(3)')->text());
         var_dump($item);
     }
 }

function html2bb($html = '')
{
    $html = nl2br(trim(stripslashes($html)));
    $html = tidy_repair_string($html, array('output-xhtml' => true, 'show-body-only' => true, 'doctype' => 'strict', 'drop-font-tags' => true, 'drop-proprietary-attributes' => true, 'lower-literals' => true, 'quote-ampersand' => true, 'wrap' => 0), 'utf8');
    $html = trim($html);
    $html = preg_replace('!<a(.*)href=(.+)>(.+)</a>!isU', '[url=$2]$3[/url]', $html);
    $html = preg_replace('!<a(.*)>(.+)</a>!isU', '$2', $html);
    $html = preg_replace('!<a(.*)href=(.+)></a>!isU', '[url]$2[/url]', $html);
    $html = preg_replace('!<br(.*)>!isU', '\\r\\n', $html);
    $html = str_replace('<p>', '\\r\\n', $html);
    $html = str_replace('</p>', '\\r\\n', $html);
    return htmlspecialchars($html);
}

 /**
  * Use the HTML tidy extension to use the tidy library in-process,
  * saving the overhead of spawning a new process.
  *
  * @param string $text HTML to check
  * @param bool $stderr Whether to read result from error status instead of output
  * @param int &$retval Exit code (-1 on internal error)
  * @return string|null
  */
 protected function cleanWrapped($text, $stderr = false, &$retval = null)
 {
     if ($stderr) {
         throw new Exception("\$stderr cannot be used with RaggettInternalHHVM");
     }
     $cleansource = tidy_repair_string($text, $this->config['tidyConfigFile'], 'utf8');
     if ($cleansource === false) {
         $cleansource = null;
         $retval = -1;
     } else {
         $retval = 0;
     }
     return $cleansource;
 }

 /**
  * @param Message $message
  *
  * @return Purchase[]
  */
 protected function parseData(Message $message)
 {
     $dom = new \DOMDocument();
     $dom->loadHTML(tidy_repair_string($message->getMessageBody(true)));
     $nodes = $dom->getElementsByTagName('p');
     $purchase = new Purchase\Purchase();
     $matches = [];
     preg_match('/((?:(?:[0-2]?\\d{1})|(?:[3][01]{1}))[-:\\/.](?:[0]?[1-9]|[1][012])[-:\\/.](?:(?:[1]{1}\\d{1}\\d{1}\\d{1})|(?:[2]{1}\\d{3})))(?![\\d])/', $nodes->item(2)->nodeValue, $matches);
     $purchase->date = new Date(join('-', array_reverse(explode('/', $matches[1]))));
     preg_match('/local: (.*?),/', str_replace('.', ',', str_replace("\n", ' ', $nodes->item(2)->nodeValue)), $matches);
     $purchase->place = trim($matches[1]);
     preg_match('/R\\$ (.*), no dia/', $nodes->item(2)->nodeValue, $matches);
     $purchase->amount = (new Currency($matches[1]))->getValue();
     return [$purchase];
 }

 /**
  * @author Hannes Gassert <hannes at mediagonal dot ch>
  * @param        string         text to be filtered
  */
 function filter($text)
 {
     /// Configuration for tidy. Feel free to tune for your needs, e.g. to allow
     /// proprietary markup.
     $tidyoptions = array('output-xhtml' => true, 'show-body-only' => true, 'tidy-mark' => false, 'drop-proprietary-attributes' => true, 'drop-font-tags' => true, 'drop-empty-paras' => true, 'indent' => true, 'quiet' => true);
     /// Do a quick check using strpos to avoid unnecessary work
     if (strpos($text, '<') === false) {
         return $text;
     }
     /// If enabled: run tidy over the entire string
     if (function_exists('tidy_repair_string')) {
         $text = tidy_repair_string($text, $tidyoptions, 'utf8');
     }
     return $text;
 }

function html2bb($html = '')
{
    $old = $html;
    $html = trim(stripslashes($html));
    $html = tidy_repair_string($html, array('output-xhtml' => true, 'show-body-only' => true, 'doctype' => 'strict', 'drop-font-tags' => true, 'drop-proprietary-attributes' => true, 'lower-literals' => true, 'quote-ampersand' => true, 'wrap' => 0), 'utf8');
    $html = trim($html);
    $html = preg_replace('!<a(.*)href=(.+)>(.+)</a>!isU', '[url=$2]$3[/url]', $html);
    $html = preg_replace('!(&lt;|<)a(.*)href=(.+)(&gt;|>)(.+)(&lt;|<)/a(&gt;|>)!isU', '[url=$3]$5[/url]', $html);
    $html = preg_replace('!<a(.*)>(.+)</a>!isU', '$2', $html);
    $html = preg_replace('!<a(.*)href=(.+)></a>!isU', '[url]$2[/url]', $html);
    $html = preg_replace('!(&lt;|<)br(.*)(&gt;|>)!isU', "\r\n", $html);
    $html = str_replace('<p>', "\r\n", $html);
    $html = str_replace('</p>', "", $html);
    return $html;
}

 /**
  * Main-Method for the Task
  *
  * @return  void
  * @throws  BuildException
  */
 public function main()
 {
     // check supplied attributes
     $this->checkDir();
     $this->checkFile();
     // load file
     $xml = simplexml_load_file($this->file);
     // Just set them
     $xml = $this->setLanguages($xml);
     // write the new xml to the old xml file, using tidy if installed
     if (function_exists('tidy_repair_string')) {
         file_put_contents($this->file, tidy_repair_string($xml->asXML(), array('output-xml' => true, 'input-xml' => true, 'indent' => true, 'indent-spaces' => 4, 'wrap' => 0, 'vertical-space' => true, 'output-bom' => false, 'newline' => 'LF', 'char-encoding' => 'utf8')));
     } else {
         $xml->asXML($this->file);
     }
 }

function tidyRepair($article, $easyset)
{
    if ('com_content' != JRequest::getVar('option')) {
        return;
    }
    if (!$article->introtext && !$article->fulltext) {
        return;
    }
    if (function_exists('tidy_repair_string')) {
        //Tidy Repair Text
        $TidyConfig = array('indent' => TRUE, 'output-xhtml' => true, 'show-body-only' => true, 'wrap' => false);
        $article->introtext = tidy_repair_string($article->introtext, $TidyConfig, 'utf8');
        $article->fulltext = tidy_repair_string($article->fulltext, $TidyConfig, 'utf8');
    } else {
        require_once 'closeTags.php';
        $article->introtext = closetags($article->introtext);
        $article->fulltext = closetags($article->fulltext);
    }
}